Biology subjects

Lindenmayer, J.

Publications and source records attributed to Lindenmayer, J..

3 recordsLinked to original sources

bioRxiv2026

ViTAMIn-O: Democratizing computer vision-based machine learning for stem cell research

Deep Learning (DL) holds exciting potential in automating the prediction of organoid differentiation results. Nevertheless, current models lack adaptability, openness, and robustness in performance. Additionally, broad employments of predictive models in wet-lab settings necessitate machine learning expertise, often not readily available in biologically oriented laboratories. To offer an intuitive solution, we present ColabViTAMIn-O, a code-free platform together with ViTAMIn-O. ViTAMIn-O is a fully open organoid-specific DL model trained and tested on a total of 34 organoid categories, incorporating annotated images across transmitted light microscopy (TLM) modalities at single-organoid resolution. It is adaptable to downstream prediction tasks of varying dataset sizes and outperforms established models even with linear-probing. It performs reliably within a few-shot framework and is even extensible to human embryo TLM imaging data at single specimen level. By releasing our platform, centralized model hub, and datasets, we hope to encourage broader deployments of specialized DL models in stem-cell laboratories.

bioinformatics↗

bioRxiv2025

Tonic interferons defend against respiratory viruses in primary human lung organoid-derived air-liquid interface cultures

Innate defences of the respiratory epithelium are the first barrier against incoming respiratory viruses. To understand the contribution of both basal (tonic) and induced interferon (IFN) to antiviral defences in a physiologically relevant system, we established air-liquid interface (ALI) cultures of primary human bronchial epithelium (HBE) and small airway epithelium (HSE). Via an organoid intermediate stage, the limited healthy donor material was expanded while preserving stemness and subsequently differentiated. Characterisation by spatial and transcriptomic analyses showed that the cellular diversity and architecture of our ALI cultures were comparable to native human lung epithelium. Upon infection with relevant human respiratory pathogens, such as Human Rhinovirus (HRV16) and human Coronaviruses (229E and NL63), only HRV16 induced a strong and early type I and III IFN response, leading to its eventual clearance from the cultures. Depletion of tonic type I/III IFNs using neutralising antibodies or scavengers reduced expression of levels of IFN-stimulated genes and increased infectious HRV production by [~]7-10-fold. Taken together, we present a method for generating primary lung epithelial cultures that retain their IFN status, demonstrate clearance of HRV by innate defences, and highlight the importance of tonic IFN in early antiviral defences. IMPORTANCEMild respiratory viral infections, for example, with human common cold coronaviruses or rhinoviruses, are a massive cause of human morbidity. The respiratory tract is the primary entry route for these viruses and also the contact site for initial innate immune defences. Here, we show that primary human lung epithelial cell-derived air-liquid interface cultures mimic the architecture and cell composition of native human lung epithelium, and retain both induced and tonic interferon (IFN) responses. Notably, our data show that the models innate immune defences are sufficient to clear human Rhinovirus (HRV) infections, which are characterised by rapid and robust IFN responses. Finally, depletion of tonic IFNs led to a marked increase in HRV infection. Thus, our research suggests that tonic low levels of IFNs contribute to the epithelial defence against viruses, maintaining the tissues immune readiness. Failure to maintain these tonic IFN levels increases the susceptibility towards infections.

microbiology↗

bioRxiv2025

Pancreatic cancer patient-derived organoids capture therapy response and tumor evolution

Patient-derived organoids (PDOs) reflect parental tumor features and may represent promising avatars for prognosticating drug response. Here, we recruited 169 patients with pancreatic cancer (PC) and established a living biobank including 83 pharmacotyped PDOs isolated from primary and metastatic, treatment-naive and pretreated PCs. In a core facility setting, the pharmacotyping success rate was 61.5%, with an unmet turnaround time of 32 days. Forty-six patients who underwent a total of 94 therapeutic lines were analyzed, resulting in a pharmacotyping-patient response matching rate of 73.4%. Sensitivity, specificity, positive and negative predictive values were 85.0%, 64.8%, 64.2%, and 85.4%, respectively. Tracing clonal evolution in longitudinal biopsies uncovered therapy-induced genetic alterations and single-nucleus multiomics identified transcriptomic and epigenetic changes associated with abnormal FGF signaling during treatment in one particular tracked study case. Our findings highlight the potential of PDOs as robust tools for drug response prediction and patient modeling to advance functional precision medicine.

cancer biology↗